Vehicle number plate recognition

For municipal traffic planning, reliable information on the utilization of roads, cycle paths, car parks, etc. is essential. For this reason, so-called cross-sectional censuses have been carried out regularly in Nuremberg since the 1970s. To this day, such censuses are largely manual work and are often carried out by students. Conventional censuses enable differentiation according to place, time, and type of vehicle, but they are not able to track traffic flows beyond the measuring point.

Often, however, it is not so much the traffic volumes at the inflow and outflow points of a complex traffic junction that are of interest, but rather the traffic flow from A to B. If it were possible to identify vehicles recorded in A and B with each other, the traffic flow between the two locations could be reconstructed from the selective measurements. It is difficult for students at the roadside to note down the number plates of passing cars, not to mention data encryption. Cameras and computer programs are needed here,

but cameras do not produce letters and numbers, only pixel images. From these, the useful information must first be extracted using image processing and pattern recognition algorithms. For photos of parked cars, software was developed within the scope of the application project described here. With its help, the duration of publicly parked vehicles can be statistically recorded, which is an important basis for the control and management of public parking space.

Freescale Cup

As part of the “Freescale Cup” educational research project, student project groups were given the opportunity to participate in  the student competition of the same name organized by Freescale Semiconductor, one of the world’s largest manufacturers of semiconductor products. The competition focuses on the field of autonomous driving: the students modify and program special model vehicle kits and use the resulting vehicles to compete against other students from all around the world. The model vehicles must be able to autonomously complete a circuit which is peppered with crossings, hills, and other obstacles.

Using a camera, the cars follow markings on the race track, which must not be left while driving. In the “Freescale Cup”, the lap time determines success and failure. For the teaching research project, the competition rules allowed teams of three students. The team took part in the Central European qualifying round, but unfortunately was unable to qualify for the European final.

Innovation Challenge

As part of the Innovation Challenge educational research project, student project groups were given the opportunity to participate in  the student competition of the same name organized by Freescale Semiconductor, one of the world’s largest manufacturers of semiconductor products. The competition focuses on the field of autonomous driving: the students modify and program special model vehicle kits and use the resulting vehicles to compete against other students from all around the world. The model vehicles must be able to autonomously complete a circuit which is peppered with crossings, hills, and other obstacles.

Using a camera, the cars follow markings on the race track, which must not be left while driving. This year's Innovation Challenge, which was held for the first time, focused on developing a prototype for a new competition concept, in which not only the lap times but also the energy consumption of the systems is crucial for the evaluation. For the teaching research project, the competition rules allowed teams of three students. The team took part in the Central European qualifying round in Deggendorf as well as in the world finals in Erlangen. In the Innovation Challenge, however, the teams took part outside the competition.